Considering these guiding principles, the S&OP process should be set up in three building blocks: scope, S&OP process, and planning of demand & supply.
1) Scope
The scope is the framework for the overall S&OP process. At first, it must be decided which hierarchies (levels) should be involved in the S&OP process and which decisions are taken at which hierarchy level. In many large companies, the starting point is country commercial, which usually owns the country demand forecast including market assumptions. Furthermore, demand scenarios are created at this level. At the regional level, assumptions and certain events are reviewed and in-region demand/supply issues discussed, e.g., shared SKUs or local supply. At the global level, demand is matched with supply from global sites. The outcome should be a consensus for a joint supply plan. Furthermore, in case of demand scenarios, supply options need to be calculated and evaluated in a so-called supply impact assessment. Topics such as end-to-end supply chain optimization and supply risk mitigations also get discussed at the global level.
Another topic that needs to be defined in the building block “scope” is the value chain coverage. Within the S&OP process, only a part of the value chain is taken into consideration as certain areas are decoupled from planning. The reason is that they are assumed to be infinite (e.g., raw material procurement) or that the information is too granular (e.g., customer of the customer).
2) Process
Once the scope is set and it is clear what kind of decisions need to be taken at specific hierarchy levels, a detailed process design follows. Hereby the S&OP decision meetings are planned backwards from final decisions to start, to cover all necessary tasks. For each meeting, an agenda needs to be created and the relevant stakeholder roles have to be defined. The RACI framework can be used to define roles and responsibilities. Preparation rounds or reviews are needed to prepare necessary decisions for the main meetings. As a rule of thumb: Every S&OP session requires one preparation round. To avoid the pitfall of overloading an S&OP meeting, the duration should not be longer than 60 minutes.
Once the decision meetings are settled, S&OP ownership needs to be decided for each meeting. The owner of the meeting is responsible for sending out invitations in time, making sure the necessary reports and decision tasks are created, and for running the S&OP meeting. Especially on a country level, we have seen people from a variety of departments running the meetings (commercials, finance, marketing, regional supply chain, etc.). It is not important which department runs the meeting, but it needs to be ensured that the owner runs it with a cross-functional perspective.
The final design choice regarding the process is the clock speed or the time interval between meetings. A monthly cycle is rather common in many organizations. There might be cases, e.g., long lead times, that indicate that a quarterly cycle is sufficient.
3) Planning
Once the decision needs are clear and the meetings have been scheduled, the content for the meeting must be prepared. Detailed planning levels and logic for demand planning, supply planning and reporting of inventories need to be established. For example, in high-margin pharma areas, effort is invested to come up with an accurate patient-built model on SKU basis to help identify parallel trade from the very beginning. In other industries, e.g., steel building capacity, utilization and grade margins are more important, so a lot of calculations are put into the supply product mix.
Within the planning horizon the reach of the demand forecast and supply plan are determined. A key driver to decide on the horizon are lead times from the value chain coverage decoupling point towards the point of demand. If data is too uncertain to predict future demand in the long run, an aggregation by product groups or larger time intervals (quarterly instead of monthly) will help to come to a good forecast.
Finally, to steer S&OP decisions, KPIs and related accountabilities are needed. It is decided what metrics are tracked and who is responsible for the creation of metrics. We strongly recommend sticking with a few standard KPIs to make monitoring as easy as possible. From our experience, two or three metrics per hierarchy level will tell 80 percent of the story. In demand planning for example, we think forecast bias and accuracy should be sufficient to evaluate forecast performance and track it over time. Target setting for KPIs should also follow a clear and reasonable guideline. Cornerstones for the target are the company’s capabilities, e.g., IT support and the environment in which it operates. For example, a bias should never be accepted, but in the case of high demand uncertainty it is unreasonable to expect high forecast accuracy.
